Identifying Evidence-Based Strategies in a Digital Mental Health Intervention for Depression: Qualitative Content Analysis

Background: Depression is one of the leading causes of disability worldwide. Cognitive behavioral therapy (CBT) is an effective treatment, but it is difficult to access due to clinician shortages, waitlists, and logistical barriers. Smartphone-based CBT interventions offer a scalable alternative to traditional face-to-face care, but few provide transparency regarding how closely they adhere to evidence-based therapeutic principles. Understanding what therapeutic components are included in interventions helps clinicians and patients determine whether they follow CBT principles and how they might help reduce depressive symptoms. Objective: This study aimed to characterize the therapeutic content of Mindset (Koa Health), a therapist-guided smartphone intervention for depression, by identifying the core CBT techniques it delivers and the specific behavioral strategies the app uses to put those techniques into practice. Methods: A qualitative content analysis was conducted on all 393 unique intervention pages of Mindset. Using established CBT strategy definitions and the behavior change technique (BCT) Taxonomy version 1 (BCTTv1), coders independently evaluated each page using a collaborative consensus approach. Interrater agreement was 93.75% for CBT and 93.62% for BCT coding. Descriptive statistics (frequency, mean, and SD) and overlap between the two were calculated. Results: All 16 core CBT techniques were identified. CBT techniques were used a total of 528 times (mean per module 66.0, SD 56.0). The most frequently used techniques included psychoeducation (164/325, 50.5% of pages), skill building (110/325, 33.8%), cognitive restructuring (46/325, 14.2%), activity scheduling (42/325, 12.9%), and self-monitoring (39/325, 12%). Across modules, 37 of 93 possible BCTs were coded 878 times (mean per module 109.8, SD 92.0) across 13 of 16 BCTTv1 categories. The most frequently applied BCT categories were shaping knowledge (205/325, 63.1% of pages), repetition and substitution (138/325, 42.5%), and feedback and monitoring (113/325, 34.8%). Overlap between the 2 frameworks was common, with the most frequent CBT-BCT pairings being psychoeducation (CBT technique)×Shaping knowledge (BCT category; appearing together on 119 pages), skill building×Shaping knowledge (80 pages), activity scheduling×Shaping knowledge (42 pages), and activity scheduling×Repetition and substitution (42 pages). Conclusions: Mindset demonstrates coverage of CBT techniques and alignment with evidence-based BCTs. This study is the first to introduce mechanism mapping, a dual-coding approach that describes the presence of therapeutic strategies and how they are behaviorally operationalized, addressing a gap in digital mental health transparency. Unlike existing content evaluations that use presence or absence checklists, our framework captures implementation depth through systematic documentation of behavioral scaffolding. This replicable methodology enables researchers to evaluate therapeutic fidelity, supports clinicians in making evidence-informed recommendations for digital mental health treatments, and provides a foundation for the development of adaptive interventions that can enhance real-world treatment outcomes for individuals with depression.